The formula
expectancy = (hit rate × average win) − (miss rate × average loss)
Every term comes from your own record. Hit rate is wins divided by total trades. Average win and average loss are exactly what they sound like, taken across every trade, not the memorable ones.
A positive number means the method makes money across a large sample. A negative number means it does not, and no amount of position sizing, discipline or conviction repairs that. Sizing decides how fast you arrive; expectancy decides where.
Why hit rate alone tells you nothing
Here is the whole argument in one chart: the reward-to-risk ratio a method needs simply to break even, at each hit rate.
Being right often and being profitable are different achievements. Only one of them is available in your brokerage statement.
Thinking in R
The most useful habit in this whole subject is to stop counting rupees and start counting R — multiples of the amount you risked on that trade.
Risk ₹5,000 and make ₹10,000: that is +2R. Risk ₹5,000 and lose it: −1R. Because your position sizing already made the rupee risk constant, R makes every trade directly comparable regardless of instrument, price or size.
Expectancy then becomes a single clean figure — this method returns 0.3R per trade — and the useful projection follows immediately: 100 trades at 0.3R and 1% risk is roughly a 30% gain before costs. That sentence contains more information than any number of win-rate screenshots.
How many trades before you believe it
Fewer than thirty trades tells you almost nothing. Random chance produces impressive-looking runs easily at small samples, which is why every new method looks brilliant for its first fortnight.
Thirty is a hint. A hundred is a reasonable working estimate. And the estimate is only valid if the trades were taken the same way — changing the rules halfway means you have two samples of fifty, not one of a hundred.
The corollary bears stating: you cannot judge a method by its last five trades. Not when they were good, and not when they were bad. Most abandoned strategies were discarded during an ordinary losing streak, which is the one thing a positive-expectancy method guarantees you will meet.
Computing expectancy from the trades you remember. Memory keeps the dramatic ones and quietly deletes the small chopping losses, which flatters every input at once. Expectancy is only meaningful from a written record of every trade, including the ones you would rather not count.
Read it yourself
Use the expectancy calculator to check yourself if you like.
- 35 wins in 100. Average win ₹6,000, average loss ₹2,500. What is the expectancy per trade?
- A method wins 68% of the time, averaging ₹1,800 per win and ₹4,600 per loss. Is it worth trading?
- You have taken 12 trades and 9 were winners. What does that tell you about your expectancy?
Show answers
1. (0.35 × 6,000) − (0.65 × 2,500) = 2,100 − 1,625 = +₹475 per trade. Wrong nearly two times in three and comfortably profitable.
2. (0.68 × 1,800) − (0.32 × 4,600) = 1,224 − 1,472 = −₹248 per trade. It loses money while winning two out of three, and it will feel successful the entire time it is failing.
3. Essentially nothing. Twelve trades is far too small a sample; a coin produces runs like that regularly. The honest answer is that you do not yet know your expectancy, and the most expensive mistake available at this point is to increase size because the early results look encouraging.
Terms used
- Expectancy
- Average profit or loss per trade across a large sample.
- Hit rate
- Proportion of trades that make money. Meaningless on its own.
- Reward to risk
- Average win divided by average loss.
- R
- One unit of risk. A +2R trade made twice what it risked.
- Sample size
- How many trades the estimate rests on. Under thirty, it rests on very little.